[Dave Haynie]

There is generally no such thing as “hardware RAID” on a motherboard. This is very often misunderstood.

RAID is very much a software protocol. When people say “hardware RAID”, they usually mean there’s software running on some other thing than the main PC’s CPU(s)… your extra hardware is a SPARC/MIPS/ARM chip running the RAID software on your RAID card, or perhaps on the CPU in an external box like my Drobo.

RAID on a motherboard is usually still running on your host CPU, just at the BIOS level, rather than at the filesystem level. “Firmware” and “software” are the same thing… we engineers tend to call it “Firmware” if it’s in a ROM somewhere. But that just means it’s always available on your MB… it’s no different in practice than software loaded from hard disk into RAM (well, technically it is, but not in a good way… Flash ROM access is hundreds of times slower than DRAM access, but the BIOS driver is only used at boot time, once you’re in Windows, there’s a Windows driver in RAM doing the same job).

So yeah, your motherboard has a Flash ROM, and that may contain a BIOS-level RAID driver, but the code from that ROM is still executed by your host CPU, just as any other code. And you probably have to load up a RAID driver under Windows to access this, which is just another chunk of driver-level code doing the RAID thing, rather than the Windows file system doing the RAID thing.

For example, a common chip on i7 motherboards is Intel’s ICH10R, which “everyone knows” does RAID.. even RAID 5. Your BIOS will set it up, and in Windows, you have to install the Intel Matrix Storage Driver to access it. But all that RAID code is running on your PC’s CPU (well, at least one of them), not offloaded to some other CPU. It’s “software RAID”, or what Linux people call “fakeRAID”.

Is it any better than using Microsoft’s NTFS RAID? Maybe. In older systems, probably not. You used to have a separate SATA or PATA “RAID” controller, which lived on the PCI bus, versus SATA build into the host chipset. So your RAID actually ran much slower and used more system resources than the built-in SATA.

These days, yeah, probably, you’re better off with the low-level driver. The SATA controller used for RAID is either in the main system I/O chip, on a private high speed chip to chip bus, or on ePCI… either way, it’s just as good as any other SATA in your system. Software-wise, even though it’s running on your host CPU, it may have small optimizations that improve performance. Intel’s driver knows very specific things about Intel’s SATA controller chip. So it may be more efficient than Microsoft’s file system approach.

Why does everyone think this is “hardware” RAID? Largely because it’s built-in to the BIOS, and in particularly, because they really don’t understand RAID very well. There is virtually no extra overhead for RAID 0 or RAID 1. And these days… benchmark a single SATA drive sometime, and see what impact that has on CPU performance. Nothing much, if it’s even still measurable (I stopped bothering, back in the PATA-UDMA days, once HDD overhead dropped below 5% CPU). Doubling that isn’t noticable. RAID 5 will have more overhead… but consider, back in the old dark days of PIO PATA, people were perfectly happy seeing 20-50% overhead from hard drive access. If Intel’s RAID 5 pushes it even as high as 5%, no one’s likely to complain.

-Dave

[Dave Haynie]

Here’s the thing.. there are different ways to create a stripeset (RAID0), or other kinds of RAID.

You set it up using the software RAID support in your BIOS. This abstracts the drive as a single drive to anything that lives above… like Windows. So you only see the one drive.

You could have skipped the BIOS setup. Windows would have seen both drives in Disk Manager, and perhaps allowed you set up a stripeset there (what’s allowed varies by the version of Windows you run). In this case, the low-level drivers see two separate drives, but the Windows filesystem manages the RAID, so at the desktop level, you see the two disks.

And of course, you can also use a “hardware” RAID, which is code for saying some other computer is doing the many->one abstraction, not the CPU(s) on your PC itself… all RAID is actually done in software. For example, I have a Drobo unit here, with four 2TB drives, which shows up over Firewire800 as a single hard drive volume. And in fact, the volume shows up as 16TB, even though there’s less actual storage there. If I were to run low, I could swap in some 3TB drives (one at a time, to allow the new drive to be used) and boost up my actual storage.

-Dave

[Dave Haynie]

Start up the Disk Manager (Administrative Tools/Computer Management/Storage) and see if you can find the stripeset listed there. From the disc manager, you can create a partition, format, etc.

-Dave

[Dave Haynie]

Blu-ray supports exactly one 3D format: MVC (Multiview Video Coding). MVC is based on AVC; it encodes one “normal” video channel and one “difference” video channel. So in essence, you get a full left-eye view, then a track of just the differences on the right… so it compresses much better. The player itself creates two full frames, one for each eye, from the Blu-ray format.

HDTVs and HDMI 1.4, on the other hand, support at least 7 different formats. Just as when HDMI devices hand-shake on resolution, your BD player and your TV will presumably handshake on 3D format. Since your BD player knows the disc is 3D, the TV reacts automatically.

The fact that you have to set the TV manually for the AVCHD disc implies that the BD player has absolutely no knowledge of the fact you have a 3D disc there. That pretty much limits the format to 3D side-by-side, half resolution, since that’s the only format the BD player would understand that’s fully transparent to the AVCHD/BD format — same principle Panasonic uses in their new 3D camcorder. It’s just a regular camcorder doing normal HD, and the 3D aspect is all interpretation of the video shot with the 3D lens attachment.

The 3D formats include field alternative, frame alternative, line alternative, side-by-side half, side-by-side full, L + depth, and L + depth + graphics + graphics depth. Only side-by-side half is transparent .. you could play that on a non-3D BD player and the 3D TV, set manually to that mode, will work. Field, frame, or line alternate send twice as many field/frames.. the video looks like 1080/48p or 720/120p, etc. A regular Blu-ray player would not have a clue about how to play such a video disc, in BD or AVCHD. Similarly, side-by-side full is practically the same as frame/field alternate, only you’re sending what appears to be a 3840×1080/24p or 2048×720/60p video… double the normal pixel rate, no non-3D TV would have any idea what to do with it. The L+ modes are even more clever.. that’s sending essentially what’s on the BD itself … left view plus difference, as I understand it. You let the TV assemble the final frames, rather than the BD player, that’s all.

The frame alternative format, also called “Frame Packing”, is mandatory on all 3D TVs that meet the SMPTE standards. All televisions are required to accept 720/50p, 720/60p, and and 1080/24p in this format, which again, is really much like 720/120p or 1080/48p… twice the video, twice the data rate, two complete images in full HD. All other formats are optional.

That’s the salient point here. Yeah, you can encode half-rez side-by-side and play it on any TV that conforms, but it’s lower resolution, and isn’t guaranteed to play. Real Blu-Ray 3D is guaranteed, because the player has all of the information, and can deliver whatever format the TV requires. I don’t believe you can do full 3D side-by-side … the BD player wouldn’t likely accept it. That’s twice the data… more overhead than a regular 3D BD, and not something a BD player is ever expecting to play. It would have to understand that format… the half-rez, it doesn’t have to… that’s a pass-through. I believe you when you say the software says it does this, but I don’t believe the software… particularly if your 3D BD player doesn’t recognize the disc as 3D.

FujiW3 encoding from Fujifilm’s W3 digital still camera. It has a 720p-ish video mode, which encode RealD side-by-side video in an AVI. This is a format you might want to input to a DVD/BD authoring program. A 3DTV with full side-by-side support could play this, but you could not encode that onto a legal Blu-ray as is… you need the MVC format.

In short, I agree that Roxio’s stuff (very consumery.. my Mom uses it) is the first consumer 3D disc creation system I’ve seen anywhere. Of course, when I got into video disc creation, we all made Super Video CDs, because VideoCD sucked and DVD wasn’t out yet. So it’s not a huge surprise that there’s some monkey business going on here… they are not delivering standardized 3D discs that you might expect to be able to play on any player 10 year hence. And while I’m sure they accept full HD or half HD or W3 as input, I seriously doubt they can produce any output format other than side-by-side half. Any other format requires your BD player to understand that format… and as you’ve pointed out, it doesn’t.

-Dave

[Dave Haynie]

Sorry for another ‘render’ question, but I just got Vegas 10 Platinum and I am so confused by the multitudes of rendering options. I don’t see one that specifies ‘GoPro’!

The render is for the thing to want. You already have GoPro video (or, actually, Cineform video… it’s no long GoPro native video). What you want is Blu-ray… so you need to hunt for an acceptable Blu-ray format.

And yeah, there are many. Vegas does Blu-ray video in MPEG-2 or AVC via the pre-defined templates, and there are several video rates and resolutions supported by Blu-ray as well. The choice of the output depends on how long you video is, how much time you have to render, etc.

The two main choices, as John explained, are Main Concept MPEG-2 or Sony AVC. The AVC format, when well encoded, looks about the same as an MPEG-2 video taking twice the space… but it takes much, much longer to create. I often use MPEG-2 for videos under 1.5 hours, AVC for larger videos.. but there’s no reason you couldn’t use AVC for everything (Blu-ray also support VC-1, but Vegas/DVDA do not… VC-1 is in-between AVC and MPEG-2 in coding efficiency).

So, your GoPro records in 1080/30p or 720/60p. If you shot in 1080/30p, you can’t really render 1080/30p for Blu-ray, since that’s not a supported format. The usual solution is to render at 1080/60i. But if you’re really shooting sports video, you probably have it set to 720/60p — that’s what I use when I shoot soccer video (different cameras, but 60p is usually what you want for action). You can render 720/60p to Blu-ray in either AVC or MPEG-2. I have templates for this… don’t recall if I wrote them or they came along with Vegas, but it’s easy enough to create your own. The basic rule is to take another Blu-ray template, the closest you can find, and then tweak that just thing that’s different.

-Dave

[Dave Haynie]

You’re talking Apples and Oranges here a little.

Your JVC Prosumer camcorder is recording in MPEG-2 long GOP at 35Mb/s, which is roughly equivalent to AVC at 17.5Mb/s, with a modern encoder anyway. I’m not sure you could convince Vegas to SmartRender to MPEG-2, but you might give it a try. Things get a little weird, because JVC uses the .mp4 or .mov (Quicktime) wrappers on these (assuming you have something like the HM100), rather than the usual MPEG-2 transport stream. But there’s MPEG-2 inside. The Main Concept MPEG-2 CODEC will render up to 80Mb/s, depending on settings (I have a high quality 80Mb/s MPEG-2 rendering right now… I’ve been using this for intermediate video, until Sony gets Cineform working under Vegas 10 — always crashes on output). Blu-ray supports slightly over 40Mb/s, so you could use your 35Mb/s if you wanted to, assuming DVDA will have it. This wouldn’t be practical for very large videos, and unless you can SmartRender, you probably will get the same quality with AVC at much lower rates. Though I’m not entirely happy with Sony’s limitations on AVC bitrate anyway; they’re fairly artificial, there’s no reason the encoder should have such limits.

Bit-rate is a little like CPU clock speed… it’s one indicator of performance, but only within the same family. Just as MPEG-2 is many times more complex, and thus, of a much higher coding efficiency than DV or Motion JPEG, AVC is many times more complex and about double the coding efficiency of MPEG-2.

On-camera AVC had been a work-in-progress until the 2009-2010 cameras. The fact is, the more complex algorithm was still maturing (and probably still is, only much more slowly), and you needed a better compression processor than many of the early AVC camcorders had available. IMHO, the quality of AVC surpassed that of HDV (25Mb/s with reduced resolution) in the major implementations over the last two years. Some companies, like JVC and Canon, have boosted the bitrate of their MPEG-2 to compete… also allowing them to use less computational juice on-board the camcorder, and drop right into existing MPEG-2 workflows (more of an issue for corporate/broadcast video than indie stuff).

Nice thing about those cameras, too, is that they record audio in 16-bit LPCM (eg, uncompressed), while HDV and AVCHD-based camcorders all use audio compression.

-Dave

[Dave Haynie]

Not that there’s anything wrong with buying a new PC, but if your prospective laptop is a potential downgrade, that won’t help your video work. Given the pretty good free anti-virus tools available, not to mention some excellent paid-for tools (I run ESET’s NOD32 on my main PC), there’s really no excuse for getting a virus. Adware can be pretty well cleansed using tools like AdAware or Spybot, both free as well.

And if you have an install disc, you can wipe just about everything by doing a clean install of Windows. You’ll have to re-install your programs as well, and there’s time involved, but shouldn’t be cost. Most fairly modern Windows PCs have a built-in means to repair or restore Windows.

My kids have a spare PC, ran me $8.00 at a yard sale, because the former owner found themselves loaded with virus and adware programs, and the local Big Box wanted over $100 to fix it. So they bought a new computer, and put this one up for cheap (only really useful for internet anyway, 800MHz Pentium 3 or some such, but hey, for less than two “Starbucks”, not too bad).

-Dave

[Dave Haynie]

Weird.. and not a general problem. At least, I’m one “Windows Defender” download away from up-to-date. Just for grins, try re-installing Vegas. Don’t know what you’re using; I’m on Vegas 10b (with the Sony AVC from 10a).

-Dave

[Dave Haynie]

You can get a BIOS-level display with just DRAM, CPU, and Video Card installed.

Fans aren’t a huge indicator, other than you have some power present. While they can be controlled via the SMA bus (system management bus), they usually default to “on” for safety, particularly your CPU and GPU fans. They also may run from different power supplies than your CPU or GPU.

One common error for the inexperienced system integrator on modern boards… you probably need at least three power supply connections.
If you bought a new power supply, it’s almost certainly an ATX12V version 2.3 supply. This will have a main 24-pin power connector for the motherboard; that’s the main supply, but not the only one to get things going.

Then you need the special additional supply lead for the CPU itself. This is a 4-pin or 8-pin connector (sometimes dual 4-pin). The connector on your motherboard is at the top left, just behind some external connectors (ATX?). Yours has an 8-pin socket, with a knock-out blocking one 4-pin connector. Only one 4-pin power cable (+12Vdc x 2, GND x 2) is needed by the CPU; the 8-pin version is for dual processor motherboard, but if your PSU has a dual connector, it doesn’t hurt to use it (I connected both on my AMDx6 motherboard).

And finally, you need a 6-pin or 8-pin ePCI power cable run to your graphics board, in most cases. Leseee… lucky you… yours want TWO 6-pin ePCI power cables connected.

So that’s the most obvious stuff. Make sure you have your three DIMMs in the appropriate color-coordinated slots (all blue or all black), and obviously, seated well.

Another thing you can do .. hook up the case speaker, if you have one. Or go find one, and hook it up. If you’re booting at all (which you should, assuming the CPU isn’t rotated — nearly impossible — and it’s getting power along with the motherboard), you’ll get a code of beeps if the MB determines something’s wrong. This kicks in early in POST… it can even tell you if the memory’s bad. Can’t tell you necessarily what’s bad about it, but it can give you a thumbs down if something’s wrong (hopefully, the printed manual that ALWAYS comes with the motherboard will have an index of error codes).

Check this all out, write back, I’ll check again sometime soon (currently in the 11th hour of a 19 hour video render…).

-Dave

[Dave Haynie]

Curious stuff… I looked it up. The RealD they mention there… same company, but a different technology. In the theater, RealD uses passive, circularly polarized glasses to deliver the proper image to each eye. In a conventional theater, this is generally via a DLP projector and a polarization wheel in front of it, but IMAX (which sometimes uses RealD, and sometimes other technologies) typically uses two projectors, one for each eye.

The “home” RealD format is, however, a “side-by-side” format… half the resolution, much the same idea as the video you get from the Panasonic SDT750. And, despite this “side-by-side” format having existed for over 100 years (“stereographs” were a fad in 1900, actually went commercial in 1849, and of course, there was the “Viewmaster” I had when I was a little kid), RealD seems to have some kind of patent on it. But I digress…

So yeah… apparently they’ve added some tag to the AVCHD format to indicate a side-by-side recording. This is NOT, however, how 3D is stored on standard Blu-ray video (MVC format). You can get players for computers that support shutter-glasses 3D on your PC using this format, too. This doesn’t actually require a 3D Blu-ray player, either, but without one that understands this other 3D format on AVCHD, you have to tell your TV to go into side-by-side 3D mode… assuming it understands that.

-Dave